A common problem in the manufacture of maleic anhydride is that it may exhibit a slight coloration from impurities in very small amounts. Even a slight color is highly undesirable for many uses of maleic anhydride, such as the manufacture of polyester resins. The specifications accepted by the industry for many purposes include an APHA color reading of less than 40 after heating for 24 hours at 140.degree. C.
Prior to the present invention, it has been known to stabilize the color of maleic anhydride with certain sulfide compounds (U.S. Pat. No. 3,564,022), certain thiophosphates (U.S. Pat. No. 3,636,057), and certain chelating agents, such as EDTA (U.S. Pat. No. 3,115,503).
Thiodipropionitrile has been used to stabilize methylchloroform (U.S. Pat. No. 3,265,747), to inhibit polymerization of conjugated dienes (U.S. Pat. No. 3,523,141), and to stabilize polybutadiene (U.S. Pat. No. 3,355,421).
Thiodipropionic acid and its esters have been proposed for stabilizing maleic anhydride.
Color-stable maleic anhydride has also been disclosed as achievable by passing molten maleic anhydride through a bed of alkali sulfates or halides (U.S. Pat. No. 3,622,600).
None of the above methods has completely solved the problem. The industry is still looking for a consistent method of assuring stability of clear color in maleic anhydride.
I have studied the maleic heat-color problem and found that it may be caused by two factors, (1) alkali metal cations or tertiary amines and/or (2) oxidation. For example, maleic anhydride is so sensitive to alkali metal ions that heat-color tests cannot be run in soft glass tubes. The small amount of sodium ions present in soft glass is enough to severely darken maleic anhydride in the 24-hour heat-color test. Maleic anhydride is also very sensitive to tertiary amines. Less than 10 parts per billion nitrogen as pyridine can produce an unacceptable 24-hour heat color. I have found that these harmful substances can readily be removed by passing the molten maleic anhydride through an anhydrous acidic cation-exchange resin. For this treatment, I prefer "Amberlyst 15", manufactured by Rohm and Haas Co. It has a "macroreticular" structure which is highly porous, and in contrast to conventional ion exchange resins, it does not lose its porosity when the swelling solvent, water, is removed. The resin is made from a sulfonated styrene divinylbenzene copolymer. Although the anhydrous resin can be used without treatment, I prefer to wash it first with a suitable solvent (hot methanol is preferred) to remove superfluous low-molecular-weight sulfonated materials because these substances can contribute to high heat color. After washing, the resin is dried and the molten maleic anhydride may be passed over it, i.e., through a bed or column. The resin-column temperatures may range from slightly above the melting point of maleic anhydride to 100.degree. C. Contact times may be from about 1 minute to about 15 minutes. The longer contact times and higher temperatures are generally not as efficient as the shorter contact times and lower temperatures, because higher temperatures and longer contact times apparently leach minute amounts of undesirable materials from the cation-exchange resin and result in high heat colors. I prefer to use about 70.degree. C. and about 2 to 10 minute contact time with the resin. At lower temperatures, longer contact time is required but at higher temperatures shorter contact time should be used. Below the melting point of maleic anhydride (52.5 C), a solvent may be used to keep the maleic anhydride in a liquid form for treatment with the cation-exchange resin if such a procedure is to be used.
Among the solvents which may be used are xylenes and other aromatic hydrocarbons that are unreactive toward maleic anhydride.
The 24-hour heat color of maleic anhydride samples originally containing large amounts of basic material has been significantly improved by treatment with only the cation-exchange resin. Other samples of maleic anhydride containing oxidizable materials as the main impurity were only slightly improved after cation-exchange treatment. But, treatment with 3,3'-thiodipropionitrile significantly improved these latter samples.
Since most newly-manufactured maleic anhydride apparently contains some of both types of impurities, the combined treatment, cation-exchange resin followed by antioxidant, yields better heat colors than either treatment separately.
In the screening of chemicals and antioxidants, it was found that many conventional chemicals normally considered to be good antioxidants were ineffective and in some cases even detrimental to the 24-hour heat color.
Antioxidants or reducing agents that improved maleic anhydride heat color were:
1. Paraformaldehyde and formaldehyde derivatives such as methylal PA1 2. Aluminum metal PA1 3. Dimyristyl thiodipropionate and thiodipropionic acid PA1 4. 3,3'-Thiodipropionitrile PA1 5. 2,5-Dihydrothiophene-1,1-dioxide PA1 6. o-Phenyl phenol PA1 7. 4,4'-Sulfonyl diphenol PA1 8. 0-Hydroxyacetophenone PA1 9. Di-n-butyl carbonate PA1 10. Zinc metal
3,3'-Thiodipropionitrile is the antioxidant I employ in my invention. It was evaluated and found to be quite effective over the range from 100 to 2,000 parts per million (ppm). It would not be practical to use it at rates over 2,000 ppm. It was still very effective at 100 ppm, is beneficial at concentrations as low as 10 ppm, and effective in proportionately lesser degrees in amounts less than 10 ppm, to as little as 1 ppm.
By far the most effective treatment, however, is the combined treatment of maleic anhydride using the strongly acidic cation-exchange resin and the addition of 3,3'-thiodipropionitrile to the maleic anhydride.
The effects of various treatments of maleic anhydride may be seen in the following Tables I-VII.
In each case, the addition of the antioxidant was accomplished by directly adding the antioxidant to molten maleic anhydride. The residence time in the column was set by the flow rate through the column. The color test used was a standard 24-hour test at 140.degree. C. in "Pyrex" color tubes.
While I prefer to use "Amberlyst 15" as the acidic cation-exchange resin, I may employ any cation-exchange resin in the acidic form; preferably the resin will be one which is readily dried and does not unduly shrink or crack in the absence of water. Another cation-exchange resin which may be used in the acidic form is Dowex MSC-1.
Solvents which may be used in the column are xylenes and other aromatic hydrocarbons that are inert to maleic anhydride.
TABLE I ______________________________________ Effect of Pyridine on Maleic Anhydride Heat Color 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 1. Untreated control (very good maleic anhydride) APHA 60 2. Control + 0.04 ppm pyridine (7 ppb nitrogen) APHA 150 3. Control + 0.55 ppm pyridine (0.1 ppm nitrogen) APHA &gt;500 4. Control + 2.7 ppm pyridine (0.5 ppm nitrogen) Brown ______________________________________
TABLE II ______________________________________ Effect of Column Treatment Only on Maleic Anhydride 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 5. Untreated control (poor quality apparently containing basic impurities) APHA &gt;500 6. Control put through Amberlyst 15 columns then heat treated APHA 60 7. Sample identical to #4 above put through column then heat treated APHA 75 ______________________________________
TABLE III ______________________________________ Effect of Column Residence Time at 70.degree. C. on Maleic Anhydride Heat Color 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 8. Untreated control (good maleic anhydride) APHA 65 9. 2 minutes residence time in column APHA 60 10. 4 minutes residence time APHA 60 11. 10 minutes APHA 50 12. 15 minutes APHA 70 ______________________________________
TABLE IV ______________________________________ Effect of Column Temperature on Maleic Anhydride Heat Color at 5 Minutes Residence Time in Column 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 13. 60.degree. C. column temperature APHA 75 14. 70.degree. C. APHA 70 15. 80.degree. C. APHA 80 16. 90.degree. C. APHA 150 ______________________________________
TABLE V ______________________________________ Effect of Antioxidant Treatments Alone on Maleic Anhydride Heat Color (1000 ppm Antioxidants) 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 17. Untreated control (fair quality) APHA 125 18. Paraformaldehyde APHA 100 19. Zinc metal APHA 100 20. Dimyristyl thiodipropionate APHA 75 21. Methylal APHA 80 22. 3,3'-Thiodipropionitrile APHA 50 23. o-Phenylphenol APHA 100 24. Aluminum metal APHA 100 25. 2,5-Dihydrothiophene-1,1-dioxide APHA 75 26. 4,4'-Sulfonyl diphenol APHA 85 27. o-Hydroxyacetophenone APHA 85 28. Di-n-butyl carbonate APHA 90 29. Thiodipropionic acid APHA 70 ______________________________________
TABLE VI ______________________________________ Effect of Antioxidant Only on Maleic Anhydride that Contains Pyridine (Note that in test #7 the ion exchange column was able to remove pyridine and give good heat color) 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 30. Control + 2.7 ppm pyridine treated with dimyristyl thiodipropionate before heat treatment Brown ______________________________________
TABLE VII ______________________________________ Combination Column and Antioxidant Treatment of Maleic Anhydride Versus Single Treatments 24 hr, 140.degree. C. Heat Color Heat Color ______________________________________ 31. Untreated control A (poor quality due to oxidizable impurities) APHA &gt;500 32. Column treatment only APHA 500 33. Antioxidant treatment only (dimyristyl thiodipropionate) 200 ppm APHA 65 34. Antioxidant treatment only (3,3' thiodipropionitrile) 200 ppm APHA 60 35. Column + antioxidant (dimyristyl) thiodipropionate) 200 ppm APHA 40 36. Column + antioxidant (3,3' APHA 40 thiodipropionitrile) 200 ppm 37. Untreated control B (good quality apparently contaning very little APHA 90 basic impurities) 38. Column treatment only APHA 125* 39. Column + antioxidant (3,3' APHA 30* thiodipropionitrile) 200 ppm 40. Antioxidant only (3,3' APHA 50* thiodipropionitrile) 200 ppm 41. Untreated control C APHA &gt;500 42. Column treatment only APHA 300 43. Column + 150 ppm 3,3' APHA &lt;20 thiodipropionitrile 44. Untreated control D (good quality with oxidizable materials as main impurities) APHA 100 45. 1000 ppm 3,3'-thiodipropionitrile added APHA 50 46. 500 ppm 3,3'-thiodipropionitrile added APHA 50 47. 200 ppm 3,3'-thiodipropionitrile added APHA 50 48. 100 ppm 3,3'-thiodipropionitrile added APHA 50 ______________________________________ *Column treatment alone was slightly detrimental yet combined treatment was better than individual treatments.